Traditionally, OSA patients are classified according to their AHI, scored during a nocturnal PSG study. This simple classification does not take into account variables such as the predominant posture for apnea–hypopnea events, the influence of sleep stage on AHI, and additional factors such as age, gender, and BMI.

In the first large study, including 574 consecutive OSA patients diagnosed by polysomnography (PSG) in a sleep disorders unit, it was found that 55.9 % were positional [8]. Thus, it became evident that more than half of the OSA patients seeking treatment are positional patients (PP). The PP were leaner and younger and had fewer and less severe breathing abnormalities as compared to the NPP. Thus, it was not surprising that they enjoyed a better sleep quality (they slept better than NPP in the lateral postures), and according to Multiple Sleep Latency Test, they were less sleepy during daytime than NPP. The respiratory disturbance index (RDI) was the most dominant parameter that predicted the positional dependency followed by BMI and age. The authors also showed that the prevalence of PP was much higher in the milder forms of the disease (mild–moderate). In these OSA patients, the prevalence of PP ranged from 65 to 69 %. The portion of patients with positional dependency decreased with increasing BMI, while age played only a minor role.

In a study on a larger population of 2,077 adult OSA patients by the same group, it was shown that 54 % were PP [9]. In this study, 60.4 % and 56.2 % were PP in mild and moderate OSA, respectively. Other authors have published similar findings [10, 11].

Furthermore, during the last years, several studies of different Asian populations have showed that the prevalence of positional OSA patients is even higher than in Caucasians. In a study of 1,170 OSA patients from Korea who underwent PSG [12], nearly 75 % of the study population were PP. However, in mild and moderated OSA, the prevalence increased to 87 % and 84.2 %, respectively. In another study on 263 OSA patients from Thailand [13], the prevalence of PP reached almost 70 %, and in a study evaluating the effect of positional dependency on outcomes of treatment with mandibular advancement device (MAD), 80 out of a total of 100 Korean OSA patients showed positional dependency [14]. Thus, the majority of Asian as well as non-Asian OSA patients are positional.

Since patients with mild-to-moderate OSA are the prevailing group of OSA patients [15], it can be concluded from the above that positional OSA comprises the vast majority of OSA patients. Thus, if PT will turn out to be a successful mode of therapy for sleep apnea patients, it will eventually be used by a large number of patients.

The high prevalence of positional OSA in the less severe forms of OSA is important, because mild OSA patients are less likely to succeed with the treatment by continuous positive airway pressure (CPAP) [16] and therefore might be good candidates for positional therapy.

The low rate of success of surgical interventions for snoring and in particular for OSA [17] and the fact that untreated patients with mild OSA might eventually progress from moderate to severe OSA [18] are additional reasons why PT should be advocated to the milder form of OSA.

Moreover, several studies performed during the last years have indicated that the worsening effect of the supine posture during sleep is valid also for patients with central sleep apnea. The first report on this issue described a patient with cerebrovascular accident (CVA) in whom a clear-cut detrimental effect on breathing abnormalities in the supine posture during sleep was found [19]. Several studies have shown that the supine posture has an aggravating effect on central sleep apnea in patients with heart failure, with or without Cheyne–Stokes breathing [20–22]. Although PT may be a new and valuable treatment tool for this particular group of patients, there is a need for additional research on this topic.

The effect of gravity on the upper airway (UA) when adopting the supine posture is most probably the dominant factor responsible for the anatomical and physiological changes in th UA observed in this posture [23]. The effect of gravity on breathing during sleep was nicely shown by Elliot et al. [24] They have analyzed 77 PSG obtained from five healthy astronauts either at normal or microgravity and found that both the severity of AHI and the severity of snoring significantly decreased in zero gravity conditions.

Several studies have suggested that sleep and posture may work synergistically to compromise the UA in OSA patients, and by adopting the supine position, one major physiological change occurs, i.e., an increase in the UA resistance [25, 26]. As a consequence, breathing during sleep becomes more labored, leading to an increased rate of episodes of partial or complete airway obstruction. The critical pressure at which the pharynx collapses (Pcrit) is an objective measurement of upper airway collapsibility. The fact that in comparison to the supine posture, Pcrit decreases in the lateral posture indicates that changing body posture from supine to lateral significantly decreases passive pharyngeal collapsibility. This fact provides a physiological support for positional therapy [27, 28].

The establishment of the concept of PP and NNP was followed by attempts to explain the anatomical basis for this concept. Pevernagie et al. have studied with fast CT scanning the size and shape of the UA in 6 PP and 5 NPP while awake [29]. The main anatomical differences between the two types of patients were present at the velopharyngeal segment of the UA. The minimal cross-sectional area (MCA) of the velopharynx for the PP group was almost twice that of the NPP group in both the supine and right lateral positions. Moreover, the shape of the tract was elliptical (with the long axis oriented laterally) in the PP group while it was circular in the NPP group. The differences in shape were predominantly due to the significantly greater lateral diameter in the PP group while the anteroposterior (AP) dimensions were similar in both groups. These data suggest that during sleep in the supine position, both PP and NPP will have breathing abnormalities mainly due to the gravity effect on the UA soft tissues which reduces the AP diameter significantly. However, when the PP group adopts the lateral position, the AP diameter is increased, and since the lateral walls are far enough apart, sufficient airway space is preserved to avoid the complete collapse of the UA. In the NPP group, changing to the lateral position will still generate pharyngeal collapse since the lateral diameter in these patients is not sufficient enough to allow significant airflow via the collapsed pharynx. Walsh et al. [30] have corroborated and extended the abovementioned concepts by studying 11 males with OSA and 11 age- and BMI-matched control subjects during wakefulness with anatomical optical coherence tomography.

In a more recent study [31], the 3D morphological features of the UA tract in positional and non-positional OSA patients matched for BMI, age, and AHI were studied by pharyngeal CT and MRI. It was found that the dominant determinant for positional dependence was the volume of the lateral pharyngeal wall. The authors noted that positional OSA patients had a smaller volume of the lateral pharyngeal wall soft tissues. Therefore, when OSA patients sleep in the lateral posture, it is the width of the lateral pharyngeal wall that will determine if the gravitational forces acting on the pharyngeal lumen are sufficient to keep it open, as it occurs in positional patients, or insufficient, as it occurs in non-positional patients, to prevent the collapse of the pharyngeal space during sleep.

The above studies provide an anatomical basis for the phenotype distinction between positional (PP) and non-positional (NPP) OSA patients. In addition, it offers an anatomo-physiological explanation for the beneficial effect of assuming the lateral position in patients with sleep apnea.

It should be mentioned that there are at least two additional factors which appear to play an important role in breathing function and are affected by positional changes during sleep:

Lung volume decreases during sleep, and when assuming the supine posture, this decrease produces a narrower upper airway which will further increases UA resistance [32]. Although the influence of assuming the lateral position during sleep was studied only in normal subjects, it should be noted that when they turn from the lateral to the supine position, the expiratory reserve volume decreases [33], and the functional residual capacity [34] and the dynamic lung compliance [35] are increased.

The genioglossus muscle is the main dilator of the UA. As such, its activity is increased in OSA patients, perhaps, in an effort to compensate for the narrower UA, decreased lung volume, and subsequent increased Pcrit [36]. It was also found that the activity of the genioglossus is decreased in the lateral as compared to the supine position in normal subjects as well as in OSA patients [37].

Although the vestibular system could theoretically play a role in increasing the activity of the genioglossus due to its role in maintaining posture, it was suggested that the postural-related changes in the activity of the genioglossus are secondary to the displacement of the hyoid bone when moving from the sitting to supine and from supine to lateral recumbent position [38].

The effect of changes in body posture during sleep on breathing function is not limited to the changes from supine to the lateral posture. For example, some OSA patients claim that the prone posture is a much better position during sleep than the supine posture. Only few studies have provided data on the effect of the prone posture on breathing abnormalities (perhaps mainly due to the fact that to sleep in the prone posture with the equipment used during PSG is almost impossible) during sleep [39, 40].

Nevertheless, a number studies have investigated the effect of elevated posture and knee up posture and the effect of head extension on breathing abnormalities during sleep.

In order to increase the likelihood of success with any treatment modality, it is essential to identify the group of patients who will benefit mostly from this particular treatment. With this in mind, we need to be able to recognize easily what are the main characteristics that will determine if a particular patient is or is not the suitable candidate for this specific treatment. This issue is of importance also for the successful outcome of positional therapy.